you need to use a 16 ohm resistor as a secondary load, and couple to the speaker through a large (20,000uf or larger) capacitor to get the DC off the voice coil. unless you have a speaker that has ferrofluid cooling, the voice coil will quickly overheat and burn out unless you isolate the DC from the voice coil. you should also have some kind of bias adjustment for the transistor bases, so you can set the DC level at the output capacitor to 1/2 the rail voltage (6V in the case of a 12V rail). also, the output power here is dependent on the rail voltage and getting the output to swing as close to "rail-to-rail" as possible. the biggest problem you are going to have with the transistors working against a resistor is a huge amount of even harmonic distortion. a better collector load to balance against would be a constant current source also running in class A. the amount of standing current will be determined by the required output power. i'm not a fan of class A amps, but if you want more power and less distortion, and to not fry your speakers, you will need to get a bit more complexity here... you probably should also have a gain and buffer stage before the output transistors.